Definition
A linear connection that assigns to each vector field a directional derivative operator acting on tensor fields, compatible with tensor type and satisfying the Leibniz rule while compensating for changes of local bases.
Principle
Principle
Extend the concept of directional differentiation to geometric objects on a smooth manifold by introducing correction terms (connection coefficients) that restore tensorial transformation under coordinate changes.
Demonstration
Demonstration
On a Riemannian smooth manifold with Levi–Civita connection ∇, the covariant derivative of a vector field V along W is ∇_W V = W^i(∂_i V^j + Γ^j_{ik}V^k)∂_j, where Γ are the Christoffel symbols determined by the metric.
Misapplication
Misapplication
Using ordinary partial derivatives in curved coordinates to compare tensor components at different points without adding connection terms; this yields non-tensorial results that depend on the coordinate choice.
Consequence
Consequence
Covariant differentiation defines parallel transport, geodesics (via ∇_dotγ dotγ = 0), and curvature tensors; it provides the mechanism to formulate intrinsic differential equations on curved spaces.
Reversal
Reversal
The ordinary directional derivative in a global Cartesian frame coincides with the covariant derivative with vanishing connection coefficients; in that flat coordinate choice no correction is required.
Boundary
Boundary
Requires a smooth vector bundle and a specified connection; not every derivative-like operator on sections is covariant—Leibniz compatibility and linearity over smooth functions fix the form of admissible connections.
Semantic Tension
Semantic Tension
Often compared with the Lie derivative: both measure change along flows, but the Lie derivative is intrinsic to flows and does not depend on a chosen connection, while the covariant derivative depends on a connection and respects tensorial contraction rules.
Synthesis
Synthesis
A covariant derivative is the connection-based rule for differentiating tensorial objects along directions on a smooth manifold that preserves tensor type and encodes parallel transport and curvature.